Answers are grounded in this paper and cite their sources.
A simple Monte Carlo model is used to simulate the avalanche process in a multiplication region which incorporates a heterojunction, intended to introduce localization into the ionization process and reduce excess avalanche noise. The results are compared with those of models where the ionization path length distribution is represented by an exponential decay, displaced from the origin by a ballistic dead space. While the latter results depend sensitively on the arbitrary choice of scheme used to evaluate the model parameters, they agree remarkably well with Monte Carlo, considering the simplicity of the model.
No takes yet. Share an insight, caveat, or question.
Groves et al. (2004) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: